ULTRA-SHORT WAVE PROPAGATION 143 



some 25 and 6 meters above sea level and under this condition the 

 effect of earth curvature cannot be neglected in the calculation except 

 for paths less than a kilometer in length. This curvature has been 

 taken into account here to the extent of replacing the curved surface 

 by a plane which is tangent to the earth at the point where the re- 

 flected ray of geometric optics touches the earth. This is justified for 

 short optical paths but cannot be used at the longer distances when 

 the receiver nearly disappears from the view of the transmitter. 



Fig. 12 shows the theoretical curve for vertical polarization based 

 on a conductivity of 1.5 X IQ-^^ e.m.u. and a dielectric constant of 

 80 e.s.u. Other values of conductivity give the same type of curve 

 but the best fit to the experimental data is obtained by this curve. 

 The dielectric constant was chosen equal to that which has been found 

 to hold for fresh water throughout this frequency range.^^ 



The agreement between the experimental and theoretical curves is 

 reasonably satisfactory. By varying one constant, the conductivity, 

 it has been possible to check approximately the absolute attenuation 

 at two distances and two frequencies. 



The conductivity (1.5 X IQ-^O is lower than that measured for sea 

 water at low frequencies. For this reason, it was considered desirable 

 to check the low frequency conductivity in this part of the bay since 

 it may have been reduced by the fresh water emptied into the bay by 

 numerous nearby streams. A number of samples were taken from 

 different points between the transmitting and the receiving locations at 

 both low and high tide. The values varied between 2.9 X lO-^^ and 

 3.7 X 10-'^ e.m.u., with an average of about 3.3 X 10-". This is 

 more than twice the value of 1.5 X 10"" indicated by the optical 

 calculations. A sample of undiluted ocean water taken at the same 

 time had a conductivity of 4.3 X 10~". 



This agreement of experiment with simple optical theory does not 

 prove that the assumed picture of a direct and a reflected wave is 

 complete. It is to be pointed out that a rigorous solution (as opposed 

 to the simple reflection picture), might require an appreciably different 

 conductivity. Mr. C. B. Feldman of these Laboratories has made 

 some short distance experiments over smooth land.-" Using fre- 



1' Since the writing of this paper, an article by R. T. Lattey and W. G. Davies 

 on " The Influence of Electrolvtes on the Dielectric Constant of Water" has appeared 

 {Phil. Mag., 12, 1111-1136, Dec. 1931). Their results indicate that the dielectric 

 constant is materially increased by salt in the water. Their experirnents were made 

 for solutions that were very much more dilute than sea water. This, together with 

 the fact that the effect of a combination of solutions was not determined, makes it 

 impossible to estimate the dielectric constant of sea water from their results with a 

 reasonable degree of certainty. . 



20 A paper covering this work will appear later: "The Optical Behavior of the 

 Ground for Short Radio Waves," C. B. Feldman. 



